Automated adaptation strategies for stream learning
Authors: Bakirov, R., Fay, D. and Gabrys, B.
Journal: Machine Learning
Volume: 110
Issue: 6
Pages: 1429-1462
eISSN: 1573-0565
ISSN: 0885-6125
DOI: 10.1007/s10994-021-05992-x
Abstract:Automation of machine learning model development is increasingly becoming an established research area. While automated model selection and automated data pre-processing have been studied in depth, there is, however, a gap concerning automated model adaptation strategies when multiple strategies are available. Manually developing an adaptation strategy can be time consuming and costly. In this paper we address this issue by proposing the use of flexible adaptive mechanism deployment for automated development of adaptation strategies. Experimental results after using the proposed strategies with five adaptive algorithms on 36 datasets confirm their viability. These strategies achieve better or comparable performance to the custom adaptation strategies and the repeated deployment of any single adaptive mechanism.
https://eprints.bournemouth.ac.uk/35640/
Source: Scopus
Automated adaptation strategies for stream learning.
Authors: Bakirov, R., Fay, D. and Gabrys, B.
Journal: Mach. Learn.
Volume: 110
Pages: 1429-1462
https://eprints.bournemouth.ac.uk/35640/
Source: DBLP
Automated adaptation strategies for stream learning
Authors: Bakirov, R., Fay, D. and Gabrys, B.
Journal: Machine Learning
Volume: 110
Pages: 1429-1462
ISSN: 0885-6125
Abstract:Automation of machine learning model development is increasingly becoming an established research area. While automated model selection and automated data pre-processing have been studied in depth, there is, however, a gap concerning automated model adaptation strategies when multiple strategies are available. Manually developing an adaptation strategy can be time consuming and costly. In this paper we address this issue by proposing the use of flexible adaptive mechanism deployment for automated development of adaptation strategies. Experimental results after using the proposed strategies with five adaptive algorithms on 36 datasets confirm their viability. These strategies achieve better or comparable performance to the custom adaptation strategies and the repeated deployment of any single adaptive mechanism.
https://eprints.bournemouth.ac.uk/35640/
Source: BURO EPrints